Improved FLIR Simulation

[JazonXD]

I think this is what you're talking about.

 

That's realistic. Try walking around your house looking through a soda straw.

 

The inaccuracies of targets marked from far away is also realistic. angular position is best at about 35-65 degrees depression angle rom any altitude. Anything above or below that and the accuracy falls off in a tangent curve looking manner (asymptotically and infinitesimally worse until the system tus off and calls it zero). Speed and the natural wobbles of the aircraft carrying the sensor has an effect as well. You mark a target instantaneously. In some sensors and systems we teach the user to mark a target over a 10 second average to tighten the mark. That ins't how the A-10C system works. You tap that TMS button and a mark appears. You can't average.

 

Also to consider is that GPS isn't what sets the position. The GPS only corrects the INS position - in most platforms at 1Hz. The position can drift a bit within that 1 second time span.

 

Further, the position of the mark is based on DTED data that is preloaded into the system. Depending on the fidelity of that DTED data, that mark can be off by a bit. Lookup DTED to learn more.

 

Keep in mind this all has a larger accuracy window than you were probably expecting. Those marks are meant to get "in the ballpark". Everything after that is dependent on weapon effects and correction with lasers and other things.

 

In summary, never expect these things to be, as we say in the trade "Dead Nuts On". They do however get you remarkably close considering everything that's happening dynamically.

 

Wow, I've seen you answering another thread about some question regarding the A-10 so I'm assuming you're the dude that actually flies them (SO COOL) but I did see some TGP short and quick footage before and it seems like that targets are much easier to identify than what they are so far in DCS. From what I saw, it was TGP with black-hot imaging and the tank was blurry but pretty easy to make out from the background. Obviously, I may have to question the reliability of the video since it IS from just YouTube on some sort of documentary, is spotting targets through the TGP on FLIR really as difficult it is in DCS as in real life? :pilotfly:

[Mike77]

USMC_Trev-

 

thank you for the thoughtful, detailed reply. Bringing the real life perspective is so valuable. Great stuff!

 

Thanks again,

 

Mike

 

 

Sent from my iPhone using Tapatalk. System specs: Windows 7 Pro, EVGA X58 3x classified 3 MOBO, i7 960 @3.2Ghz, 24Gb Ram, Nvidia EVGA GTX 780ti with 3 Gb VRAM.

[HeadHunter52]

Speaking for everyone who has ever worked to locate and track a target with a sensor: You can't imagine how difficult it is in real life. Depending on the MRT or resolution of the camera you're using, plus the realities of what the atmosphere does to both EO and IR images, finding something without specific cueing from another source is next to impossible unless the target is within a mile. That's the truth of it. Welcome to "more realistic".

 

And that's why they pay me the big bucks.

 

Sure helps to have real-world experience and input. Hog drivers relating their take on things is golden.

[hansangb]

Judging by his name, I'd have to assume he's tangentially involved with A-10s. Marines don't fly A-10s. Unless he branch transferred from the AF to the Marines.

[Yurgon]

[...] finding something without specific cueing from another source is next to impossible unless the target is within a mile.

 

Thx a lot for the input!

 

Just to clarify, do you mean that it's almost impossible to find a target at more than 1 mile slant range from the aircraft? Or do you mean it's almost impossible to locate a target outside of a 1 mile search radius? The difference between these two is obviously quite enormous. ;)

 

Did you actually fly the A-10C in 1.5 with stock settings? Is that the type of realism you describe? That would be great to know! :thumbup:

[baltic_dragon]

What USMC_Trev says definitely is interesting, still after recently having read a book (which I strongly recommend everyone to check): "A-10 Thunderbolt II Units of Operation Enduring Freedom 2008 - 2014" I really doubt TGP spotting can be that hard - there are dozens of situations described by real-life A-10C pilots where they were talking how introduction of the TGP totally changed their life in terms of spotting targets.

 

And while there were also stories confirming what USMC_Trev says about limitations of the pods, most gave impression that they could get a very detailed view of the battleground. Below two short citations:

 

‘With the targeting pod we could see what looked like black or dark spots near the centre of the three vehicles. It turned out to be hydraulic fluid from the middle vehicle, which was disabled. The enemy was shooting at them from the high ground on the left, and also taking pot shots at the MRAPs from the right."

 

‘During one of my orbits I used the targeting pod to spot a group of middle-aged males who were leaving a building in a compound not too far from “Halo 07’s” position. SIGINT [signal intelligence] sources began talking about a possible Taleban C2 [command and control] location within a mile or two, and I duly confirmed this following my sighting. It was soon acknowledged from the ground that the building was indeed the C2 compound.

[Shez]

Some real Targetting pod footage. Now imagine there's trees, foliage, etc.

 

[ame]

[/ame]

[hansangb]

Thanks for that. Just ordered the book. I'd also recommend this book about an F16 pilot flying Wild Weasel missions. Very immersive in how he describes his experience.

 

 

[StandingCow]

Since I haven't done it in this thread here is two screen shots I captured while flying a night mission, one is black hot the other white. See if you can even spot the tank.

 

 

 

[JazonXD]

Since I haven't done it in this thread here is two screen shots I captured while flying a night mission, one is black hot the other white. See if you can even spot the tank.

 

~snip snip~

 

 

Show us the answer, master!

 

Please don't tell me it was a troll and there wasn't even a tank in the picture :D

[StandingCow]

Show us the answer, master!

 

Please don't tell me it was a troll and there wasn't even a tank in the picture :D

 

 

In the top image it is to the top left of the xhair, this is black hot.

 

In the bottom image it is right on top of the xhair, this is white hot.

 

Now, I have no experience with these sensors in real life, but based on the youtube videos posted... targets are not this difficult to see, they look like they did back in version 1.2. These near invisible objects have to be some sort of bug with 1.5... ED apparently are aware and investigating.

 

Edit: Looks like the thread got merged into the wishlist again... even though it isn't really a wishlist subject. 1.2 TGP was fine, 1.5 it is not.

 

Yurgon had a good 1.2 to 1.5 comparison here: http://forums.eagle.ru/showpost.php?p=2610156&postcount=31

Edited January 5, 2016 by StandingCow

[JazonXD]

In the top image it is to the top left of the xhair, this is black hot.

 

In the bottom image it is right on top of the xhair, this is white hot.

 

Now, I have no experience with these sensors in real life, but based on the youtube videos posted... targets are not this difficult to see, they look like they did back in version 1.2. These near invisible objects have to be some sort of bug with 1.5... ED apparently are aware and investigating.

 

Edit: Looks like the thread got merged into the wishlist again... even though it isn't really a wishlist subject. 1.2 TGP was fine, 1.5 it is not.

 

Yurgon had a good 1.2 to 1.5 comparison here: http://forums.eagle.ru/showpost.php?p=2610156&postcount=31

 

Wow, even after telling me where they are, I can't even spot them. I must be pretty bad at this!

[Sgt_Baker]

Finding the target is all the fun. Actually doing something to it is rather elementary once you've got that far! :)

[Boberro]

Some real Targetting pod footage. Now imagine there's trees, foliage, etc.

 

 

I do think real view is better, what I mean:

 

- video resolution, encoding ect

- we do not know original resolution

- we do not know scaling - it is easier to put even ugly image on bigger screen and spot something than otherwise

 

Example Ka-50's Shkval is really old, poor system but CRT TV seems to be not so small.

To have real view of pilot we would have to get real resolution of the displays, resolution of the "TV pod" and MFD\TV screen physical size.

[Sgt_Baker]

...resolution of the "TV pod" and MFD\TV screen physical size.

 

I've actually studied this rather extensively, as you might imagine. One of the key variables - and often the most overlooked - in a sim is the angular resolution of what you're looking at compared to real life. In this instance we're talking about in-cockpit displays, but the same applied to everything rendered outside the aircraft.

 

F.ex. I calculated that observing the A-10C TGP expanded to "full screen height" on my 24" HD displays is approximately equivalent to what you'd see in the actual cockpit in terms of real-world resolution, give or take. Furthermore, I think the TGP is presently modelled on an approximation of the Litening II/Litening G4, so no matter how large you make the export, it still renders world-imagery at a maximum of 1024x1024.

 

Having worked rather extensively with IR cameras in the past, it's true that the IR views are somewhat different (thinking mostly of 1.2.6 here) to what one would expect in real life, as they appear to be based on an image processing technique as opposed to modelling actual thermal emissions.

 

Edit: The reason a certain fruit-related manufacturer refers to some of its displays as "retina" is that the perceived resolution of those displays actually approaches the visual acuity of the MkI Eyeball, which is what I'm waffling on about here.

Edited January 5, 2016 by Sgt_Baker

[Yurgon]

Everything I'm talking about is based on real life, and on airborne sensors in general. This sim has way more variables I can't come close to addressing, like settings, video cards and their peculiarities and all that. I'm just discussing why players difficulties finding targets in DCS mirror our own real world challenges and how they are analogous.

 

Thank you, good to get some RL input! :thumbup:

 

OTOH, I still don't know if 1.5 makes thing more realistic than 1.2 or less. :music_whistling: :D

 

As I recall it, targets used to be ridiculously easy to spot in an earlier version of DCS A-10C and people with RL experience kept telling us that it's not how it looks IRL. Then at some point it was made a bit more difficult and I guess more realistic. My impression is that the change now introduced in 1.5 was an unintended side-effect of the otherwise improved textures, and judging by RL videos I think it is now too difficult in the sim, as in "much more difficult than it would be IRL under comparable circumstances".

[Chazz_BMF]

I talked to a pilot once, he told me how they identified "hidden" vehicles fx. vehicles parked under structures by looking at the tracks on the ground. Maybe ED should implement tracks, that would for sure make it easier to find the targets. Just follow the bread crumbs :)

[Yurgon]

That reminds me of a story Ed Macy wrote down in "Apache" (or might have been "Hellfire") where they spotted a group of people going into a field at night and, upon returning to their houses, leaving some type of matter behind that was still visible in the IR.

 

They later found out that this was, I believe, known as communal defecation... :D

[hansangb]

Hell, as a lowly infantrymen, my NVGs could spot footprints after a person walked across the grass. Didn't last long, but you can see it clear as day. I can't imagine what these expensive sensors/cameras can see! And this was, what , 25 years ago???? Damn..time flies.

[Bob Denny]

Hell, as a lowly infantrymen, my NVGs could spot footprints after a person walked across the grass. Didn't last long, but you can see it clear as day. I can't imagine what these expensive sensors/cameras can see! And this was, what , 25 years ago???? Damn..time flies.

Yes, I can tune the FLIR on our helicopter so that it will see tractor tire prints for example in a field 4 hours after the tractor goes through. It is amazing how tiny differences in temperature can show up.

Edited January 6, 2016 by Bob Denny

[moggel]

Hi everyone,

 

I haven't taken the time to read every post in this thread so it might be that what I'm about to say has already been dismissed, but I got here via a link that Yurgon was kind enough to give me when I was weeping over this issue in a different thread.

 

Just a few thoughts:

1. Wouldn't it make sense to separate the visual settings for FLIR (and other camera models) from the normal graphics settings?

2. Wouldn't it also make sense to apply a functional filter to model heat signatures, as opposed to setting up individual textures for vehicles/buildings etc? A functional approach would allow for a more dynamic model where the signature could be affected by idle/runnings engine, outside/ground temperature etc.

 

Just a few thoughts ...

[Bob Denny]

I think Arma has some scheme for this.

[Boberro]

Yup, Arma model seems to be really simple and cost effective...

[Yurgon]

Just a few thoughts:

1. Wouldn't it make sense to separate the visual settings for FLIR (and other camera models) from the normal graphics settings?

2. Wouldn't it also make sense to apply a functional filter to model heat signatures, as opposed to setting up individual textures for vehicles/buildings etc? A functional approach would allow for a more dynamic model where the signature could be affected by idle/runnings engine, outside/ground temperature etc.

 

All good ideas, and I think most of them have already been discussed here. :)

 

1. Would be a relatively "cheap" way, if it's possible, but we'd have to hear from ED if that even works.

 

2. Great idea, but a lot of work. Some people even volunteered to do the work for all models, but I don't think there's been any official feedback yet.

[Home Fries]

2. Wouldn't it also make sense to apply a functional filter to model heat signatures, as opposed to setting up individual textures for vehicles/buildings etc? A functional approach would allow for a more dynamic model where the signature could be affected by idle/runnings engine, outside/ground temperature etc.

2. Great idea, but a lot of work. Some people even volunteered to do the work for all models, but I don't think there's been any official feedback yet.

 

Implementing #2 could be as simple as applying a Photoshop style Gamma value or Exposure setting to the base texture while looking through a FLIR. The exposure setting could be based on the base temperature of the object (including the ground based on temperature).

 

Applying exposure or gamma is a pretty simple thing to do to a texture, and implementing it, while more complex, could still be fairly straightforward with tables and linear interpolation (e.g. X amount of gamma for a cold BTR-80, Y for one that has been moving and is hot, Z for somewhere in between because it just started up, etc.).

 

This would be a significant improvement over what we have now, and wouldn't be as much of a development time sink as a full thermal simulation.